On 05.01.2022 04:35, Forest Simmons wrote:

It's a perhaps surprising but still nice observation that it doesn't
matter how you eliminate candidates in STV as far as the DPC is
concerned: as long as you do it one at a time, the criterion still holds
because the election condition only looks at first preferences.

I like the idea of using Ranked Pairs to determine the elimination
order. Because RP passes LIIA, this gets rid of a major source of chaos
in STV/IRV because the eliminations don't change the social order. Only
the election and surplus distribution steps do that.

It may be a mar on the perfect proportionality of STV that Ranked Pairs
has a centrist bias (since it's a single-winner election method), and
thus RP-STV would have some element of centrist/consensus bias. On the
other hand, you could cast this as a feature rather than a bug: there's
certainly no lack of polarization in the current political world, so
perhaps a bit of centrist/consensus bias is not so bad.

As for Coombs STV, that's certainly one possible interpretation of
binomial STV. But if binomial STV is Coombs STV, then it definitely
isn't monotone, because in the single-winner case it's just regular old
Coombs.

-km

I'm sure it's more subtle than Coombs because Binomial STV, like
Proportional Approval Voting PAV, [according to the impression I have], is
supposed to evaluate slates of candidates before definitively eliminating
anybody.

I like the idea of RP or MAM determining a fixed Elimination order. I
wonder what Steve Eppley would say???

El mié., 5 de ene. de 2022 4:30 a. m., Kristofer Munsterhjelm <
km_elmet@t-online.de> escribió:

On 06.01.2022 04:59, Forest Simmons wrote:

Strictly speaking, it's not entirely fixed: the order will usually
change after someone is elected and the surplus is redistributed. But
it's much closer to fixed than ordinary STV!

-km

KM, Fair enough.

Where a voting methods exclusion count rates on the scales of
measurement should be some indication.

I refer to the measurement scales given by SS Stevens, in the journal,
Science: classificatory; ordinal; interval; ratio, in order of
increasing power of measurement.

Usually, exclusion counts are on a scale of more or less (ordinal scale)
to determine who is excluded.

FPTP uses a single count, that is ambiguously an election count or an
exclusion count or both, which is an ordinal scale count.

Traditional stv including Meek uses an ordinal scale exclusion count --
Last Past The Post, instead of first past the post.

Binomial stv uses both a rational election count and a rational
exclusion count. These counts are averaged, to give an over-all result,
still in terms of keep values, extending on the keep value (ratio of
quota over a candidates total vote) introduced by Brian Meek.

The binomial stv exclusion count, of voter unpreference, is the reverse
or inverse of the election count. And is inverted to give a
supplementary election count. The two counts can then be averaged with
the geometric mean, giving greater and more precise use of preferential
information than traditional stv.

Regards,

Richard Lung.

On 04/01/2022 22:55, Kristofer Munsterhjelm wrote:

Thank you, nothing to do with Mr Coombs.

Binomial stv is monotonic: it gives well behaved results. Switching
votes to a candidate can not perversely help lose them the election.
Because, the candidates keep values are re-calculated accordingly.

Switching the quota from Hare to Droop etc does not change the order of
merit of the candidates (tho technically it changes their electibility).

There is no premature exclusion (or election) of candidates. The result
depends on their over-all keep values, after all the preferences have
been counted forward (electively) and backward (exclusively).

Regards,

Richard Lung.

On 05/01/2022 12:30, Kristofer Munsterhjelm wrote:

In the SS Stevens scales of measurement, the classificatory scale is
less precise than the ordinal scale. So, in my view, it is better to
have ranked choice or preference votes than classes of votes. (I take it
that approval voting uses the latter.)

Binomial stv is a transferable vote (unlike approval voting, I believe).
Indeed binomial stv is a transferable vote of unpreference (its
exclusion count) as well as of preference (the traditional stv election).

Regards,

Rchard Lung.

On 06/01/2022 15:05, Kristofer Munsterhjelm wrote:

On 06.01.2022 20:52, Richard Lung wrote:

Okay, I think I see what you mean by different types of scales. But how
do I determine whether some election method that's provided to me (as an
algorithm) has an exclusion count at all? What does it mean for a method
to have an exclusion count?

(BTW, I know the classificatory scale as the nominal or categorical
scale. There are probably other names, too.)

-km

KM,
There is, I believe, a deep answer to your question, which is the one that occurs to me first, happily or unhappily.
That is, there is a structure of measurement, that is always with us, but we are only more or less aware of it. (We might borrow the term "deep structure" off Noam Chomsky.)
Any election system would thus be judged by how aware it is of this measurement structure. To take an example of minimal awareness, Congress gerrymandering, say "sweetheart gerrymandering", to use the technical term for a Republican and Democrat senator, who collude to make their district boundaries safe seats for each other.
Barack Obama refered to this malpractise in his closing state of the union address. It was to the effect that voters should choose their representatives, instead of representatives choosing their voters.
The point is that this gerrymandering is both an "election" and "exclusion" process - carried out by a hand-full of politicians, instead of the nation. It is still a transferable vote, not conducted by the Hare system, for the populace, but by the gerrymanderers for themselves.

In good science generally, including the Hare system, a common structure of measurement is evident. In bad science, like gerrymandering, the structure is stunted almost beyond recognition.

Regards,
Richard Lung.

PS. Yes, classificatory scale sometimes called the nominal scale.

On 6 Jan 2022, at 10:30 pm, Kristofer Munsterhjelm km_elmet@t-online.de wrote:

Okay, I think I see what you mean by different types of scales. But how
do I determine whether some election method that's provided to me (as an
algorithm) has an exclusion count at all? What does it mean for a method
to have an exclusion count?

(BTW, I know the classificatory scale as the nominal or categorical
scale. There are probably other names, too.)

-km